Rules are Made to be Broken - How Machine Learning Can Fix the SME Credit Gap

The small to medium enterprise or SME is the foundation stone of our business community. Smaller firms account for up to 95 percent, sometimes more, of the business landscape across many countries. Across the European business landscape of 2015, for example, companies with fewer than 250 employees represented 99 percent of all firms. As such, SMEs are major employers, making vital contributions to the economy of a country. In the US, small businesses have contributed almost half of the nation’s GDP and employ almost 85% of the nation’s workers. In the Asia Pacific region, we have examples like Infosys of India who from a capital input of USD$250, is now worth US$4 billion. SMEs are by their nature and by necessity competitive, agile, and innovative.

However, all is not well in the world of SMEs. It takes courage to begin your own company, and it takes vision to take it to the next level of employing people and increasing market share. To keep a small firm buoyant and working, the company needs to have a credit line and often capital injection. Current models of financing the SME are proving to be not fit for purpose. This post will explore the reasons why that is so and how innovations in technology, such as machine learning, can ensure our vital smaller enterprises are properly funded for success.

Why Are SMEs Invisible to Financial Institutions?

Organizations that fit into the definition of an SME, which in Europe is under 250 employees and in the US under 500 employees, have global markets which they can reach out to. Technology has opened the world to SME trade and a smorgasbord of opportunity is laid bare, offering engagement opportunities like never before. But to dampen this mood, a World Trade Organization report “Levelling the trading field for SMEs”, found that finance was a major stumbling block to taking advantage of market opportunities.

“Lack of, or insufficient access to, finance can strongly inhibit formal SME development and trade opportunities.” WTO

The report goes on to identify the huge impact on trade that underfunding has on SMEs; for example, smaller companies have barriers to market entry such as the prohibitive cost of shipments. Without proper funding, smaller firms can neither grow nor compete.

And, it isn’t just western companies feeling the pinch of poor funding decisions. The lack of funding into SMEs in emerging markets is even more acute. The situation is being termed the “credit gap”. This credit gap is more of an expansive chasm and clearly illustrates how the basic financial needs of an SME that are currently not being met. The gap is a staggering $1.5 trillion according to calculations by the Asian Development Bank.

Why it stands at such a high figure can be traced back to visibility.

SMEs are formidable, yet invisible. As far as credit is concerned, the SME has an invisibility cloak around their financial transactions. Public availability and timeliness of financial data are two of the biggest hurdles in decloaking an SME. The current system of scoring which is static and linear does not apply to an organization that often uses non-standard data that doesn’t fit the traditional model. Old methods such as the Altman Z-score no longer fit the bill or reflect the true record of an SME. In this new world order, we need to use new world vision to see SMEs clearly. We need to turn to the latest tools on the block to allow us to use less obvious forms of data and then use it smartly.

How Machine Learning Makes the Invisible, Visible

“AI tools provide the most compelling and straight forward way to increase returns than virtually any other technology available in the working capital space” - Tom McCabe, US Country Head of DBS Bank

One of the issues of the SME is that it has a “thin-file”. That is, smaller organizations struggle to present the depth of data needed for more traditional creditworthiness decisions. To do the SME justice, we need to address these misconceptions and create new pathways to knowledge. The Fintech space has opened up channels and models of doing business that challenges conventional thinking. The world has been introduced to new ways of banking from the challenger banks and new modes of operating are opening up new ways of working. To shore up these disruptions artificial intelligence and its subset machine learning (AI and ML) have entered, stage left. This technology allows new approaches to age-old problems, including how to solve the credit gap

Because AI and ML can utilize data from non-traditional sources, it can effectively be used to create a thick file from a thin file. AI solutions can take alternative data sources and extract the right information to make difficult decisions. The SME credit gap can be bridged at last.

“Traditional scorecard methods are static and create ‘blind spots’. The innovation that Flowcast has made around smart data analysis has a number of key benefits that remove these blind spots to give true visibility to thin-file, credit invisible SMEs.” - Ken So, Founder and CEO of Flowcast

The Digital DNA That Holds the SME Together

SMEs have the potential to become a powerhouse for success. With the right level of funding, smaller companies will bring wealth and prosperity to their directors, employees, the funding vehicle, and the country. It is vital that we support such organizations by creating financial structures that understand the, often unique, needs of SMEs. Making the invisible, visible, using artificial intelligence and machine learning will allow banks to feel confident that they have met the criteria of the risk.

The use of smart profiling that is tailored to the SME market is the answer to the SME funding problem.

Machine learning gives us the technology to build smarter systems that can utilize otherwise hidden data. Meeting the funding needs of the SME will give our banks a potentially lucrative new market and inject the economy of a country with new growth.